Did Zoroastrianism Invent Satan?

¶ Did Zoroastrianism Invent Satan?

Short answer. No, on the evidence we can actually test, and the test is a control group almost nobody runs. The usual case for Zoroastrian origin is a resemblance: the New Testament Devil looks like Angra Mainyu, so Persia is assumed to be the source. But resemblance is not a test. The test is to ask which branch of Judaism had the most exposure to Zoroastrianism and see what its adversary became. That branch is Babylonian rabbinic Judaism, which lived under Sasanian Zoroastrian rule for roughly four centuries, and its Satan got smaller: “Satan, the evil inclination, and the Angel of Death are one” (Babylonian Talmud, Bava Batra 16a). Meanwhile the branch that moved west, away from Zoroastrian rule entirely, produced “the god of this age”. Exposure and dualism run in opposite directions. What Persia demonstrably did send is an administration: a court, an office, a law and a prosecutor. The Devil was assembled later, and further west.

Tier note: the loanword measurement and the Talmudic text are bedrock, checkable by anyone against a published dataset and a Hebrew Bible. That the Achaemenid administration is the model for the heavenly court in Zechariah 3 is one scholar's argument, strong and peer-reviewed but not consensus, and it is marked here as contested-but-grounded. The inference from the two-branch comparison is a reconstruction: it is good evidence about direction, and it is not a demonstration.

¶ Resemblance is not a test

The argument almost always runs the same way. Zoroastrianism has Ahura Mazda and Angra Mainyu, a good power and a hostile one, locked in a struggle that ends with the hostile one defeated. Christianity has God and the Devil, and the same arc. The two pictures rhyme, Persia ruled Judea for two centuries, and so the conclusion is drawn.

The trouble is that a rhyme is compatible with several histories. It can mean borrowing. It can mean that two traditions facing the same problem, why the world is full of evil if God is good, reached for a similar shape independently. It can mean both traditions inherited something older. A resemblance on its own cannot tell you which, which is why comparative claims need a control: some case where the hypothesis predicts one thing and its rivals predict another.

There is such a case, it has been sitting in plain sight for a long time, and as far as we can find nobody has run it as a test. It is the subject of this page.

¶ What Persia demonstrably sent, measured

Start with what can be counted rather than compared. Every language that borrows from another leaves a fingerprint in what it borrows. Borrow cooking words and the contact was kitchens. Borrow court words and the contact was offices.

Tag the non-Semitic loanwords of the Hebrew Bible and Biblical Aramaic against an external linguists' taxonomy and the Iranian fingerprint is sharp: 39.5 percent of 38 Old Iranian loanwords fall in law and political office, against an 8.3 percent expectation if they were spread evenly (exact p = 1.6 x 10^-7), rising to 50 percent if you keep only the etymologies nobody disputes. The words are a staff directory: dat royal law, pitgam edict, 'ahasdarpan satrap, gizbar treasurer, detabar judge, 'adargazar counselor.

And the row that matters most for this page: zero of those 38 fall in the taxonomy's Religion and belief field. Not a low number. Zero. The same books borrowed cultic vocabulary from other donor languages, so the channel was open and simply was not used from this direction. The full dataset, its pre-registration and its limits are set out in The Persian Passband.

That is a measurement of a channel, not of a doctrine, and it cannot by itself settle where an idea came from. What it does establish is the register in which Persia was actually speaking to Judea: government.

¶ The satan has a job description, and it is administrative

Now read the adversary himself, and notice that in his earliest appearances he is not doing anything demonic. He is doing paperwork.

Asked where he has been, he answers that he has been “roaming the earth and going to and fro upon it” (Job 1:7, 2:2). The verb is shut, and the same verb carries “the eyes of the LORD” that range through the whole earth in 2 Chronicles 16:9 and Zechariah 4:10. To be exact, Job has the simple stem and the eyes-of-the-LORD passages have the intensive, so this is the same root in different forms rather than the same word twice. He is not wandering. He is on patrol, and he reports.

The same root names an actual piece of imperial paperwork. At Ezra 4:6, in the reign of Ahasuerus, opponents of the Jerusalem rebuilding lodge a sitnah against the inhabitants of Judah and Jerusalem: a formal written accusation submitted to the Persian court. It is built on the same root as ha-satan. The root that eventually names the Devil is, in the same corpus and under the same empire, the word for a memo of denunciation.

Jason Silverman has argued the point in detail for Zechariah 3, where the satan stands at the right hand of the high priest Joshua to accuse him. On Silverman's reading the scene is not a priestly investiture but a satrapal confirmation hearing, with the satan as the administrative accuser inside the satrap's court. The Achaemenids really did run these procedures: the Pherendates correspondence has a Persian satrap asserting his right to vet candidates for an Egyptian priesthood, having already rejected two. And accusation was a routinized office with a routinized risk, which is why the trial of Tiribazus ends with the accuser Orontes punished for accusing falsely. Zechariah 3 ends the same way: “The LORD rebuke you, O Satan”.

Silverman's own conclusion is the sentence to keep. The satan, he writes, was originally “not demonic at all, merely a functionary designed to ensure local officials were unlikely to commit treason against their overlords”, and over time “the content was retained within the tradition but the original context was lost”. A real office was absorbed, its setting was stripped, and what remained was read as something it had never been. More on the office itself in What Does Satan Mean in Hebrew?.

¶ The control group, which points the other way

Here is the test. If proximity to Zoroastrianism produces a dualist Devil, then the Jewish community with the most Zoroastrian exposure should have the most dualist Devil. So find that community.

It is not hard to find. Babylonian rabbinic Judaism lived under Sasanian Zoroastrian rule for roughly four centuries, from 224 to 651 CE. Same cities. Middle Persian vocabulary throughout the Babylonian Talmud. Documented contact with the Zoroastrian priesthood. There is an entire modern scholarly field, Irano-Talmudic studies, devoted to mapping the interaction. If anywhere, here.

What that community produced is Bava Batra 16a, where Reish Lakish teaches: “Satan, the evil inclination, and the Angel of Death are one”. Three names, one entity, and the proof text is Job 2:7, the Jobian prosecutor himself. The adversary is folded inward into the human moral faculty and the ordinary fact of dying. He is still God's. He still needs permission. Four centuries of living next to actual Zoroastrians, and the Jewish adversary got smaller, more internal and more firmly subordinate.

Now the other branch. Christianity moved west, into the Greco-Roman world, out from under Zoroastrian rule entirely, and produced an adversary who is “the god of this age” (2 Corinthians 4:4), “the ruler of this world” (John 12:31), who commands a kingdom (Matthew 12:26) and who can offer the kingdoms of the world because “it has been delivered to me” (Luke 4:6).

The exposure and the dualism run in opposite directions. That is the finding. The branch soaked in Zoroastrianism produced the least dualistic Satan in the tradition, and the branch that never lived under it produced the Devil.

The within-branch check is sharper still, because the Babylonian rabbis were not sealed off and did engage Iranian material. Sara Ronis, working on rabbinic demonology, finds that Babylonian rabbinic demonic discourse tracks older Sumerian and Akkadian content while adopting only the legal form of the Zoroastrian discourse around it. Form, not content. Which is the loanword result stated in another register, arrived at independently, by a scholar reading entirely different material.

¶ Then where was the Devil made

If the composite Enemy was not imported, he was built, and the seam is visible in one sentence: “And the great dragon was thrown down, that ancient serpent, who is called the Devil and Satan, the deceiver of the whole world” (Revelation 12:9, repeated at 20:2).

Four traditions welded in a single clause. The chaos dragon of ancient Near Eastern combat myth. The serpent of Genesis 3, who in the Hebrew Bible is never identified as Satan. The diabolos of the Greek translators. And ha-satan, the prosecutor of Job and Zechariah. Before this verse they are four things. After it they are one thing.

Revelation did not invent the identification, and this page does not claim it did. Wisdom of Solomon 2:24 and the Life of Adam and Eve already reach for it. Revelation is the canonical locus rather than the first occurrence, which is a claim about distribution and not about priority. The layered build-up is traced in Where Did Satan Come From in the Bible?, and the parallel fusion that produced hell out of four separate words in The Devil Was Assembled.

¶ What this does not show

It does not show that Iranian religion had no effect on Judaism. Persian period contact was long and real, serious scholars argue for influence on eschatology, and the general question is treated on its own terms in Did Zoroastrianism influence Judaism?. This page addresses one figure.

Vocabulary is not theology. A loanword count shows which register crossed. It cannot show that the heavenly court was redecorated, and we would be arguing against our own evidence if we pretended otherwise.

The semantic category may be anachronistic. Religion as a field separable from law and kingship is a modern European construct, and Jonathan Z. Smith spent a career making that point. Achaemenid administration and Achaemenid cult were not separate spheres. Our zero may be measuring the absence of a category rather than the presence of a filter, and we do not have an answer to that objection.

Much of the Achaemenid administrative evidence is Greek. Xenophon, Diodorus and Plutarch are the sources for a good deal of it, and Albert de Jong's warning applies at full strength: Greek reports of Persian institutions are Greek constructions. Silverman flags the problem himself and calls parts of his evidence fragmentary. His is one article, not a consensus, and he explicitly brackets the question of the later demonic development as beyond his scope.

Absence of dualism in rabbinic Judaism is not proof of absence of influence. Branches diverge for many reasons, and a community can refuse an idea it has met. The control is strong evidence about direction. It is not a demonstration, and this page does not upgrade it into one.

And the resemblance to Angra Mainyu is functional, not ontological. The New Testament Devil is created, subordinate and defeated. Christianity never became a religion of two equal principles, and anyone arguing that it did has overstated the case in the other direction.

¶ Common questions

¶ Did Zoroastrianism invent Satan?

Not on the evidence that can be tested. The case rests on a resemblance between the Christian Devil and Angra Mainyu, and resemblance cannot distinguish borrowing from parallel development. When you run the comparison as a controlled one, the result inverts: the Jewish community that lived under Zoroastrian rule for four centuries produced a smaller and more subordinate adversary, not a larger one. What Persia measurably supplied was administrative and legal vocabulary, including the courtroom setting in which the biblical satan does his work.

¶ Is the Christian Devil the same as Ahriman?

They resemble each other in function and not in metaphysics. Angra Mainyu, Ahriman in the later form, is the hostile principle opposed to Ahura Mazda from the beginning. The New Testament Devil is a created being, subordinate to God, permitted to operate for a time and certain to be defeated. Christianity is not a two-principle religion and never became one, which is why the resemblance is worth noticing and not worth overstating.

¶ If Persia did not give Judaism Satan, what did it give?

An administration. Old Iranian loanwords in the Hebrew Bible concentrate in law and political office at 39.5 percent against an 8.3 percent chance expectation, and zero of the 38 fall in the field of religion and belief. The vocabulary is satraps, treasurers, judges, counselors, edicts and royal law. The earliest satan behaves like part of that apparatus: he patrols, he reports, he requires authorization and he files formal charges.

¶ Why did Jewish Satan not become the Devil under Persian rule?

That is the question the usual account has to answer and rarely addresses. Babylonian rabbinic Judaism had the deepest and longest Zoroastrian exposure of any Jewish community, roughly four centuries of Sasanian rule, and its adversary was folded into the evil inclination and the angel of death rather than expanded into a rival god. Sara Ronis's work on rabbinic demonology finds the same pattern in miniature: the rabbis borrowed the legal form of Zoroastrian demonic discourse while their content tracked older Mesopotamian material.

¶ Is ha-satan in the Book of Job the same as the Devil?

No. In Job the word carries the definite article, which is how Hebrew marks an office rather than a name: the satan, the way you would say the prosecutor. He appears among the sons of God as a member of the divine court, he needs permission before acting, and he is never called evil. In Numbers 22:22 the angel of YHWH himself stands in Balaam's road as a satan, which is decisive: the word names a function that God's own messenger can perform.

¶ What does Bart Ehrman say about Zoroastrian influence?

He has moved away from it. Having once accepted the thesis, he wrote in 2019 that he is “not so sure. At all”, citing the difficulty of dating Zoroastrian texts earlier than the Jewish material and asking what evidence there is for Persian influence in third-century Judea. He is scrupulous about the other side, pointing readers to Anders Hultgard's essay on Persian eschatology to balance the sceptical case. This page agrees with his caution about the doctrine and offers the measurement he said he did not have for the channel.

¶ Where did the Devil actually come from, then?

From assembly rather than import. Revelation 12:9 welds four previously separate figures into one in a single clause: the chaos dragon, the serpent of Genesis 3 who is nowhere in the Hebrew Bible identified as Satan, the Greek translators' diabolos, and the Jobian satan. The identification was already being reached for in Wisdom of Solomon and the Life of Adam and Eve, so Revelation is the canonical locus rather than the first occurrence. What made it decisive was that this was the version that entered the canon.

¶ Sources

Job 1:6-12; 2:1-7; Numbers 22:22; Zechariah 3:1-2; 4:10; 2 Chronicles 16:9; Ezra 4:6; Genesis 3; Daniel 12:2; 2 Corinthians 4:4; John 12:31; Matthew 12:26; Luke 4:6; Revelation 12:9; 20:2; Wisdom of Solomon 2:24. Babylonian Talmud, Bava Batra 16a (Reish Lakish). Jason M. Silverman, “Vetting the Priest in Zechariah 3: The Satan between Divine and Achaemenid Administrations”, Journal of Hebrew Scriptures 14, art. 6 (2014), DOI 10.5508/jhs.2014.v14.a6, open access; and Persepolis and Jerusalem (2012). Sara Ronis, “A Seven-Headed Demon in the House of Study”, AJS Review. Shai Secunda, The Iranian Talmud: Reading the Bavli in Its Sasanian Context (2013); Yaakov Elman on Irano-Talmudica. Ryan E. Stokes, The Satan: How God's Executioner Became the Enemy (2019); Peggy L. Day, An Adversary in Heaven (1988); Elaine Pagels, The Origin of Satan (1995); Henry Ansgar Kelly, Satan: A Biography (2006). Albert de Jong, Traditions of the Magi (1997), on the limits of Greek reporting of Persian religion. Benjamin J. Noonan, Non-Semitic Loanwords in the Hebrew Bible (2019); Shaul Shaked, Encyclopaedia Iranica, on Iranian loanwords. Jonathan Z. Smith, Drudgery Divine (1990), on the category problem. Loanword data, method and limits: the Passband Corpus v0.2, CC BY 4.0, pre-registered at DOI 10.5281/zenodo.21879071. CC BY 4.0.

The volume behind this page: doi.org/10.5281/zenodo.21879071

¶ Related

More from the sourced library:

• Where Did Satan Come From in the Bible?
• What Does Satan Mean in Hebrew?
• The Persian Passband
• Did Zoroastrianism influence Judaism?
• Is the Devil in the Old Testament?
• The Devil Was Assembled